Digital twins are enabling utilities to better understand their systems and optimise improvements. Pilar Conejos explains what is required to develop and use a digital twin with ultimate success.
There is a growing need for more sustainable and resilient cities because of increasing population and climate change. Within this context, informed decision-making should employ advanced tools that combine real time sensor data gathering, advanced analytics, and model-based capabilities to simulate ‘what if’ scenarios. A ‘digital’ or ‘virtual’ twin of a defined system can fulfil these requirements.
The digital twin (DT) philosophy is not new. It has been used since the 1960s by the National Aeronautics and Space Administration (NASA), USA, because of the need to remotely operate and maintain systems – although the twin developed then was physical rather than virtual. As an example of its usefulness, NASA was able to successfully perform the rescue of the Apollo 13 lunar landing mission thanks to its twin on Earth.
The DT concept gained recognition in 2002 through the work of the internationally renowned expert Michael Grieves. Initially, the DT aimed to optimise the life-cycle of a product in relation to its design and manufacture process, together with the subsequent maintenance required during its lifetime. According to Grieves, a virtual twin model must contain three main parts: the physical assets, a virtual model, and the connections of data and information that tie virtual and real spaces together.
DTs are not only useful in industry; they can also be developed and exploited within a city management context, particularly in the drinking water supply system. While the majority of these systems have been implemented in the industrial sector, particularly in the manufacturing, aerospace and defence industries, there has been a notable increase in their use in the water sector.
A DT should replicate real system behaviour in a virtual model, serving as the basis for experimentation – i.e., a virtual copy of reality that allows any simulation under any condition to be performed. In the case of water systems, a complete DT would be a virtual replica of all the processes that take place from the water source to the user’s tap and, finally, to the natural environment, where the water is returned. The great advantage of a DT is the complete and holistic vision it provides, so having a DT of the integral water cycle would be the desirable final objective, enabling more efficient decision-making that takes into account all of the variables involved, knowing the overall impact of decisions across all processes. For example, a DT can demonstrate how an improvement in the level of leakage or the addition of new infrastructure would affect the quality of the service provided in terms of water pressures and water quality, network operation, overall energy consumption and raw water treatment. However, this is not as straightforward as it may seem because of the number and complexity of the processes that take place in the water cycle.
Deployment strategy
One strategy is to create a DT of each stage of the water cycle (drinking water treatment, water distribution networks, wastewater networks and wastewater treatment) or specific processes within them, with the ultimate aim of interconnecting all the DTs, considering that outputs of some DT processes can be used as inputs for others. However, it is essential that the DT is built on a solid technological base in order to ensure scalability.
A DT can be used for a variety of purposes. Indeed, it can be used to solve and address many of the problems related to water systems management, including, among others, energy optimisation, water quality improvement, water loss reduction, early anomaly detection, and decision support in emergency conditions. In any case, a DT will always enable better understanding of a system’s operation, helping to improve its operation and design to achieve the different objectives set.
Before deploying a DT it is essential to define achievement objectives. To do so, it is important that the water company’s challenges and needs are fully explored. Following this analysis the scope of the DT can be defined – i.e., the water cycle or the part of it that should be included in the DT, as well as the capabilities it should offer. It is important to highlight that DTs can evolve and grow. So, a DT may start deployment with some specific objectives (e.g., early warning systems or energy optimisation) and scope (e.g., the water distribution network or even just the transmission line system) and develop to respond to further objectives and broader scope.
Audit data and tools
The next step is to define the data required, which will depend on the scope and objectives to be covered by the DT. The type of data required can be static (related to physical assets) or dynamic (real time data). Most of the time they are located in different isolated systems within the company (e.g., Geographic Information System, Supervisory Control and Data Acquisition, Computerised Maintenance Management System) or in external systems (e.g., weather forecasts). If this is not the case, a plan should be put in place to collect the necessary data. As the DT must be built on solid foundations, it is necessary at this point to have a platform capable of integrating all the data coming from the different sources, and to have procedures in place to guarantee the life-cycle of the data, from its acquisition to its consolidation on the platform.
It is important to stress, and not to overlook, the importance of having procedures in place in the utility to ensure the collection of ‘manual’ data in systems such as GIS or CMMS, as well as the collection of automatic information, such as that sent by sensors, where the correct choice of sensor, communication system and maintenance is particularly important. It is impossible to build a DT mirror of the real system if it has been built and continuously fed with incorrect data.
The objectives to be covered by the DT and the complexity of the process to be replicated will also define the simulation model, which may be physics-based (hydraulic models), data-based or both. In whichever case, a platform must enable models to be built or imported and connected to data, ensuring that the connection is live and maintained over time. It is also important to define the information that the DT will offer, its interactions and user interface.
Throughout this process, it is important not to forget that the DT has to be adopted by people, so they have to be involved in the process from the beginning. In fact, the adoption of a DT must be part or a consequence of a digital transformation of the utility. This implies a new way of working across the utility.
The scale of the challenges for the water sector can only be overcome by a combination of technology, engineering principles, and the experience and expertise of its people. DTs are very much about people and processes, as well as technology. A DT is unique because it provides a holistic view, bringing together different technologies, data and, most importantly, promoting transparency and collaboration, breaking down silos of data and knowledge.
The author: Dr Pilar Conejos is a member of IWA’s Digital Water Programme Steering Committee. She is Digital Twin Product Manager at Idrica and was previously responsible for network control and operation for Greater Valencia at Global Omnium. She is also a part-time professor at the Universitat Politècnica de Valencia, Spain.
The author is grateful to Global Omnium and Idrica for the opportunities to put this technology and philosophy into practice.
